The AI bubble debate and ballooning valuations have dominated stock market discourse this year.

Looking at startup exits shows which kinds of companies are finding buyers, and who's making money when they do.

When a startup gets bought or goes public, shareholders may receive cash or stock in exchange for their ownership stakes. The theory goes that those payouts can also seed new businesses, fund other investments and put money back into the community.

That makes exits a potential way for capital to circulate beyond the companies that created it. 

But that doesn't happen automatically. How many people share in the cash back — and what they do with that money — can determine whether a startup's success has a ripple effect or simply adds to the wealth of those who already have it.

So far this year, AI is part of the exit picture, but the activity spans a range of industries. Healthcare companies accounted for the largest share of US startup exits during the first three quarters of this year: 64 of 170, according to PitchBook and NVCA's quarterly Venture Monitor reports. Meanwhile, 58 companies had an artificial intelligence and machine learning designation, including eight of those in healthcare.

Geographically, exit activity tracked familiar patterns. San Francisco led the country with 35, followed by Boston with 19. San Jose and New York each recorded 18. The data show where exits are happening, but not what the deals are worth or where the money goes afterward.

Specialization can be a factor in keeping wealth local. Boston's numbers show how and why that happens: 14 of its 19 of its exits were healthcare companies, reflecting the region's slew of health-related universities, research institutions, specialized labs and investors, according to Forrest Wright, an independent research analyst.

Unlike software or AI, “the type of innovation and startup activity [in Boston] can’t be ported to Silicon Valley,” Wright said.

Those connections can help keep companies, jobs and related economic activity in the region. Whether the financial returns stay depends in part on where shareholders live and how they use the proceeds.

Wright also pointed to Colorado's aerospace and defense industry. Its Air Force presence and established aerospace sector give companies in those fields resources they might not find elsewhere.

Sure, there’s a lesson in specialization, but it doesn’t guarantee exit success. 

Pittsburgh is known for robotics and applied AI, for example, and saw just two exits. Philadelphia is a center for cell and gene therapy — but its two exits were not in those fields. Baltimore’s two exits happened in a hotspot for biotech. Washington DC, with a specialization in cyber and fedtech, saw four exits.  

Startups are staying private longer, while the traditional exit routes of acquisitions and IPOs have struggled to keep pace with the industry's growth.

Companies that raised big during the 2020 and 2021 funding boom are now facing a tougher market. Some are selling for less than investors expected, leaving less money and resources flowing outward.

“They reached valuation that they definitely can't IPO at,” Wright said. “They're facing this tough dilemma of do we just try to get acquired at a much lower valuation, or do we try to kind of keep on trying to chug along here?”

❝

Another angle:

“Sharing knowledge across different industries or related industries helps regions succeed more than sharing within the same industry. Specializing in a single industry was not helpful, and measures of clustering were found to not matter much.”

researchers Georgios Fotopoulos and Haifeng Qian in Small Business Economics

↓

💼 Good jobs in cool places

Pittsburgh

North Dakota

Maryland

Remote

↓

Pick your faves: Vote now

Who's the emerging entrepreneur that made the biggest splash this year? The most meaningful launch? The best event? The attention-grabbing deal? Your chance to choose happens now: Voting is open for the annual Technical.ly Awards.

Based on reader nominations, we've curated four finalists in each of six categories across our core ecosystems. Cast your ballot at technical.ly/vote by Nov. 25.

↓

💰 Top resources for hot entrepreneurs

Awards up to $20,000 in non-dilutive seed funding to launch organizations that expand access to proven poverty-alleviation solutions in underserved communities.

Who it’s for: Aspiring social entrepreneurs launching new businesses or NGOs focused on distributing proven poverty-alleviation interventions.

Geography: Open internationally; Type: Pitch competition / Seed grant; Deadline: Nov. 15 (regular); Dec. 6 (if you register for an extension); Cost: Free.

Connects startups developing sustainable food, packaging and materials solutions with multinational corporate partners, mentors and industry experts without taking equity.

Who it’s for: Early- to growth-stage startups working in agrifood, packaging, biomaterials and circular economy technologies.

Geography: Open internationally; remote participation; Type: Accelerator / Corporate innovation challenge; Deadline: Nov. 25; Cost: Free.

Offers practical, no-code strategies for using AI to automate customer communications, create content and streamline everyday business operations.

Who it’s for: Small business owners and non-technical founders interested in integrating AI into their workflows.

Geography: National; virtual event; Type: Workshop / Webinar; Date: Nov. 12, 11 a.m. to noon CT; Cost: Free.

↓

↓

3 things to know

🍎 Just Google it: Google is giving schools more control over student access to AI, but they can't fully separate Google Search from its AI features. Districts in California and North Carolina are raising concerns, arguing that students should be able to browse the web without AI-generated answers. [EdSource/GovTech]

💼 AI CEOs: Delaware, a state known for its business-friendly policies, is considering a proposal to let AI agents run companies with little human involvement. The plan would allow approved businesses to test AI-run operations for 30 months. [TechRadar]

🙏 A higher power: A University of Kentucky employee has received religious accommodation allowing him to avoid using AI at work. The employee, who identifies as a Humanist, argued that AI conflicts with his beliefs about human creativity and responsibility. [HR Dive]